IEA Sees Data Center Power Demand Near 1,000 TWh by 2030 as AI Lifts Grid Strain

2 articles · Updated · ETEnergyWorld · Jun 9

900-1,000 TWh of electricity could be consumed by global data centers in 2030, the IEA says, up from about 460-500 TWh today and nearing 3% of world power demand.
AI-driven expansion of GPUs, servers and other digital infrastructure is the main driver, shifting competitiveness from raw computing speed toward performance-per-watt, cooling efficiency and heat management.
20% demand shares in some data-center clusters could pressure local grids even if national averages stay lower, raising risks for reliability, adequacy, decarbonization and infrastructure resilience.
Quantum computing could intensify that challenge because cooling loads at near-absolute-zero temperatures may match or exceed computing energy, making thermal engineering and cryogenic systems central to future compute planning.
India illustrates the broader policy issue: data centers are projected to rise from under 1% of electricity demand now to about 6% by 2047, requiring integrated planning across power, water, cooling and digital infrastructure.

Sources

ETEnergyWorld12h ago

IEA: Global Data Center Power Demand to Nearly Double to 1000 TWh by 2030 Due to AI

electricchoice.com12h ago

U.S. Data Center Power Map by State (2026)

As AI's thirst for power grows, can we turn data center waste heat from a liability into a valuable community resource?

Beyond AI, is quantum computing's extreme cooling demand the next great challenge for our global energy infrastructure?

Will the AI revolution inadvertently reverse climate progress by pushing power grids back towards fossil fuels?

The AI Data Center Power Surge: How Explosive Growth Threatens Grids, Drives Up Costs, and Forces a Global Energy Reckoning

Overview

Driven by the rapid rise of artificial intelligence, data center construction surged between 2024 and 2026 as tech giants pushed forward to meet growing digital demands. This expansion caused a significant increase in the energy footprint of digital infrastructure, putting immediate and substantial strain on existing power grids. As a result, energy supply strategies for critical facilities had to be reevaluated, with many developers in the US turning to onsite natural gas-based power generation. However, these onsite projects faced technical and financial challenges, highlighting the urgent need for new solutions to support the evolving demands of AI-powered data centers.

...